Long-Term Cost-Effectiveness of Fractional Flow Reserve–Based Percutaneous Coronary Intervention in Stable and Unstable Angina

Background There are limited studies on the cost-effectiveness of fractional flow reserve (FFR)-based percutaneous coronary intervention (PCI) over angiography-based PCI. Objectives The current study sought to evaluate long-term cost-effectiveness of FFR-based PCI compared to angiography-based PCI. Methods A cost-effectiveness analysis was conducted using a nationwide cohort that consisted of patients with stable or unstable angina from the National Health Insurance Service (NHIS) and Health Insurance Review and Assessment (HIRA) database in Korea. The cost-effectiveness analysis was also performed by using a decision and Markov model with key values from the United States and the United Kingdom health care systems. Incremental cost-effectiveness ratio (ICER), an indicator of incremental cost on additional quality-adjusted life-years gained by FFR-based PCI, was evaluated. Results In the NHIS-HIRA data, FFR-based PCI was used during the index PCI in 5,116 patients (3.8%) among 134,613 eligible patients. FFR-based PCI showed significantly lower risk of all-cause death (5.8% vs 7.7%, P = 0.001) and spontaneous myocardial infarction (1.6% vs 2.2%, P = 0.022) than the angiography-based PCI at 4 years. In the NHIS-HIRA data, FFR-based PCI gained 0.039 quality-adjusted life-years at a lower cost ($303) than angiography-based PCI, yielding an ICER of −$7,748 during the 4-year follow-up. FFR-based PCI was dominant in the health care system of Korea (ICER = −$7,309), United States (ICER = −$31,267), and United Kingdom (ICER = −$1,341) during a 10-year time horizon. These results were consistently shown in probabilistic sensitivity analyses. Conclusions In the current cohort, FFR-based PCI was associated with higher quality of life at a lower cost than angiography-based PCI. FFR-based PCI was cost-effective in patients with stable or unstable angina undergoing PCI.

M easuring the fractional flow reserve (FFR) during an invasive coronary angiography identifies a stenosis that causes myocardial ischemia and enables better selection of patients likely to benefit from a percutaneous coronary intervention (PCI).Previous studies confirmed that FFR was superior in guiding PCI in patients with stable ischemic heart disease than conventional invasive coronary angiography alone. 1Even though FFR is a class 1A recommendation in current guidelines for the assessment of intermediate stenosis in stable ischemic heart disease, 2 the penetration rate of FFR in contemporary practice varies across different health care systems and also remains limited due to multifactorial reasons such as physicians' attitude, knowledge barrier, and environmental barrier (reimbursement, procedural time, or medical cost). 36][7][8] Particularly, previous Kingdom with different health care systems. 10,11e Korean NHIS covers approximately 97% of Koreans, while the 3% of remaining Koreans who cannot afford national insurance are covered by the Medical Aid Program. 12Claims submitted for reim-  Hong et al Cost-Effectiveness of FFR-Based PCI  for the cost and utility of U.S. and UK based on previous literature. 13We projected the discounted lifetime health care cost by multiplying the number of subjects with the sum of the costs in every health status.QALYs were estimated using the utility values associated with each health status multiplied by the proportion of subjects living in the status.Total QALYs were accumulated from the QALYs values in each cycle.
In the U.S. and UK health care systems, the value of $60,000 and $40,000 per QALY was considered a reasonable WTP based on the GDP per capita.Multivessel Evaluation] trial). 10,11The transition Hong et al Infarction] trials). 15,16dical costs for the Korean population were estimated from individual patient-level claim data.
Regarding death-related costs, we estimated the medical costs within 6 months before death.To obtain the annual medical cost, cumulative medical costs were divided by the observation period.[26] To further assess the intraindividual and parameter uncertainties, PSA by the Monte Carlo Simulation was performed wherein subjects were randomly sampled and simulations repeated 25,000 times to obtain the outcomes.As for the input variable ranges in the simulation, lognormal distribution was used for transitional probabilities, beta distribution for utilities (utility value ranged between 0 and 1), and gamma distribution for costs (costs could not be <0).
As a sensitivity analysis, modeling with the different costs and utilities from the different health care systems (U.S. and UK) was performed using input parameters extracted from the previous evidence (Table 1).and À$6,892 for U.S. and UK, respectively (Table 5).In the PSA analysis, given the GDP per capita in each country, the likelihood iterations of cost-effectiveness for FFR-based PCIs were 93.5%, 92.3%, and 90.8% for Korea, U.S., and UK, respectively (Figure 1).
The results of the sensitivity analyses by applying 0%, 3.5%, and 4.5% for the discount rate showed Hong et al Cost-Effectiveness of FFR-Based PCI  3).In the scenario of change in the time horizon from 5 to 15 years, the ICER decreased from À$7,701/QALY to À$7,296/QALY in Korea, from À$31,912/QALY to À$30,437/QALY in the U.S., and from À$1,359/QALY to À$1,276/QALY in the UK, respectively (Supplemental Table 3).

DISCUSSION
The

GAP OF EVIDENCE AND LIMITED ADOPTION RATES
OF FFR-BASED PCI.In contemporary practice, the presence of inducible myocardial ischemia is the prerequisite for PCI, and current guidelines support the use of invasive physiologic indexes such as FFR or nonhyperemic pressure ratios for the decision of revascularization in stable ischemic heart disease as a Class Ia recommendation. 2These strong recommendations are based on multiple RCTs and large-scaled registries, which consistently demonstrated significantly better clinical outcomes following FFR-based PCIs over angiography-based PCIs. 4 Due to the limited diagnostic yield of noninvasive functional tests in distinguishing myocardial ischemia originating from epicardial coronary stenosis, invasive physiologic interrogations possess clinical relevance in patients with chronic coronary syndrome.Two recent large observational studies from nationwide cohorts further supported the survival benefit of FFRbased PCI over angiography-based PCI. 10,30In the same line, the current nationwide study also identified significantly lower risk of all-cause mortality and spontaneous MI following FFR-based PCIs in the exclusively revascularized patients for stable or unstable angina.Nevertheless, the global adoption rate of the invasive physiologic assessment with FFR is about <6%. 3 Similar to that of many other countries worldwide, the adoption rate of FFR in Korea was 3.8% in the current study.Although previous surveys revealed that there are heterogenous reasons for the limited adoption rates of FFR, physicians' attitude, knowledge barrier, and environmental barrier (reimbursement, procedural time, or medical cost) might be some of the most important reasons. 36][7] In the absence of evidence for the cost-effectiveness of FFR, the additional medical cost due to measuring FFR could act as a hurdle for the use of FFR in decision-making during PCI.

COST-EFFECTIVENESS OF FFR-BASED PCI OVER
ANGIOGRAPHY-BASED PCI.6][7][8] In the FAME trial, FFR-based PCIs in patients with a multivessel disease provided higher QALYs and lower medical cost than angiography-based PCIs. 5 In the FAME 2 trial, PCIs rather than medical treatment in patients with FFR #0.80 led to improved clinical outcomes as well as quality of life without an increase in cumulative medical costs ($16,737 AE $13,108 vs $16,792 AE $10,139, P ¼ 0.94) at 3 years. 6Although the PCI group had higher initial medical costs, the medical treatment group had higher follow-up costs which was driven by the higher rate of urgent revascularization.As a result, the ICER for PCI compared with medical treatment was $1,600.The cost-effectiveness     In the subgroup analyses, FFR was more costeffective in the high-risk patients for ischemic heart disease, such as diabetes mellitus, or in patients who share common risk factors with ischemic heart disease, such as atrial fibrillation or heart failure.These findings suggest that the benefit of appropriate lesion selection through FFR at the index procedure would be greater for high-risk patients than for low-risk patients.Furthermore, among 3 different health care systems, cost-effectiveness of FFR-based PCI was more remarkable in the U.S. health care system, which has higher medical cost than Korean or UK system.This difference in cost-effectiveness was mainly attributable to relatively lower medical costs incurred when adverse clinical events were treated in Korea or UK, whose public health care system regulates medical costs.These results emphasize the greater potential benefit of FFR in patients with multiple comorbidities and in health care systems with higher medical costs.Fifth, the current data could not represent the use of nonhyperemic pressure ratios.Finally, and importantly, this study is only able to show that FFR-PCI is cost-effective in patients selected for this strategy.
We were unable to study differences in angiographic or procedural characteristics between groups.
studies have limited generalizability as they were based on either randomized controlled trials (RCTs) with strict inclusion criteria or registry studies that could not thoroughly examine the cost or quality of life.In addition, little is known about whether costeffectiveness of FFR-based PCI varies across different health care systems.In this regard, the current study sought to evaluate the long-term cost-effectiveness of FFR-based PCI compared to that of angiography-based PCI using the National Health Insurance Service (NHIS) and Health Insurance Review and Assessment (HIRA) claim database in Korea.In addition, the current study also aimed to evaluate the generalizability of costeffectiveness of FFR-based PCI across different health care systems using the Markov model.METHODS STUDY DESIGN AND DATA.The current study was performed in 3 phases.Phase 1 was a nationwide cohort study using the NHIS and HIRA nationwide administrative claims database in Korea to generate a comparative prognosis between FFR-and angiography-based PCI. 9 Phase 2 was an individual patient-level cost-effectiveness analysis of FFR-based PCI over angiography-based PCI using phase 1 data.Phase 3 was the cost-effectiveness analysis of FFRbased PCI over angiography-based PCI based on a decision and Markov model.Key inputs in the model were acquired from phase 1 and 2 data and previous evidence from the nationwide registry as well as RCTs conducted in the United States or the United bursement to NHIS and Medical Aid Program are reviewed by the HIRA service, a central office in the Korean Ministry of Health.This study was approved by the Institutional Review Board of Samsung Medical Center, and informed consent was waived as deidentified data were used.This study followed the Consolidated Health Economic Evaluation Reporting Standards reporting guideline. 13STUDY POPULATION OF PHASE 1 AND 2. In this study, all men and women aged 18 years who underwent PCI for stable or unstable angina at secondary or tertiary hospitals between January 2011 and December 2017 (N ¼ 332,629) were included.The index PCI was defined as the first PCI that was performed during this period.To investigate the underlying comorbidities before the index PCI, patients who had medical records of <1 year prior to the index PCI were excluded from the study (N ¼ 136,522).Patients who had their index PCI for acute myocardial infarction (MI) (N ¼ 61,382), had a history of coronary artery bypass graft surgery (CABG) (N ¼ 6), or were lost to follow-up immediately after the index PCI (N ¼ 106) were also excluded.A total of 134,613 eligible patients were divided into 2 groups according to the use of FFR during the index PCI into FFR-based PCI or angiography-based PCI.Follow-up clinical data were assessed until December 31, 2018.DATA COLLECTION AND CLINICAL OUTCOMES.The NHIS-HIRA database contains information about A B B R E V I A T I O N S A N D A C R O N Y M S CI = confidence interval FFR = fractional flow reserve HR = hazard ratio ICER = incremental costeffectiveness ratio MI = myocardial infarction PCI = percutaneous coronary intervention QALY = quality-adjusted life-year WTP = willingness-to-pay Medicine, University of Iowa Carver College of Medicine, Iowa City, Iowa, USA; g Division of Cardiology, Department of Internal Medicine, Chonnam National University Hospital, Gwangju, South Korea; and the h Department of Internal Medicine and Cardiovascular Center, Chosun University Hospital, University of Chosun College of Medicine, Gwangju, South Korea.*Drs Hong and Kim contributed equally to this work.
death.Patients initially allocated to angina could either remain in that same state or transition to spontaneous MI or death in every cycle, according to transition probability (Supplemental Figure 1).Patients who were allocated to spontaneous MI could transit to death over time.Patients who developed unplanned revascularization without MI or death remained in the same state.We simulated subjects with equivalent characteristics as the trial population in phase 1 and modeled the costs and health utilities for 10 years beyond the phase 1 study.A time horizon of 10 years was applied for the model because the starting age for the simulated subjects was 66.8 years, and life expectancy in Korea is around 80 years.Cost and utility data were discounted by an annual rate of 4.5% according to the Korean Guidelines of Methodological Standards for economic evaluation and 3.5%

2 0 2 2 : 1 0 0 1 4 5
Cost-Effectiveness of FFR-Based PCI probability for death after spontaneous MI was obtained from RCTs (COMPLETE [Complete versus Culprit-Only Revascularization Strategies to Treat Multivessel Disease after Early PCI for STEMI) and FLOWER-MI [Flow Evaluation to Guide Revascularization in Multivessel ST-Elevation Myocardial current study investigated the long-term costeffectiveness of FFR-based PCI compared to angiography-based PCI in stable or unstable angina patients who underwent a PCI using the NHIS-HIRA claim database.In this large-scale nationwide data set, FFR-based PCI was shown to achieve better quality of life at lower cost than angiography-based PCI.In addition, in the model-based cost-effectiveness analysis, FFR-based PCI was consistently more CENTRAL ILLUSTRATION Cost-Effectiveness of Measuring Fractional Flow Reserve in Patients Undergoing Percutaneous Coronary Intervention Hong D, et al.JACC Adv.2022;1(5):100145.The current study investigated the long-term cost-effectiveness of FFR-based PCI compared to angiography-based PCI in stable or unstable angina patients who underwent PCI using the NHIS-HIRA Claim Database.In this large-scale nationwide data set, FFR-based PCI was shown to achieve better quality of life at a lower cost than angiography-based PCI, which was mainly driven by the lower risk of all-cause death and spontaneous MI.In addition, in a model-based cost-effectiveness analysis, FFR-based PCI was a consistently dominant treatment than angiography-based PCI among 3 different health care systems: Korea, US, and UK.FFR ¼ fractional flow reserve; ICER ¼ incremental cost-effectiveness ratio; MI ¼ myocardial infarction; NHIS-HIRA ¼ National Health Insurance Service and Health Insurance Review and Assessment; PCI ¼ percutaneous coronary intervention; QALY ¼ quality-adjusted life year.

J 2
A C C : A D V A N C E S , V O L . 1 , N O . 5 , 2 0 2 Hong et al D E C E M B E R 2 0 2 2 : 1 0 0 1 4 5 Cost-Effectiveness of FFR-Based PCI cost-effective than angiography-based PCI in 3 different health care systems: Korea, U.S., and UK.Various sensitivity analyses including PSA showed consistent results supporting the cost-effectiveness of FFR-based PCI over angiography-based PCI (Central Illustration).

FIGURE 1
FIGURE 1 Incremental Cost-Effectiveness Planes for FFR-Based PCI Compared With Angiography-Based PCI Until 10 Years After PCI Among Different Health Care Systems Cost-Effectiveness of FFR-Based PCI FFR-based PCIs across Korean, U.S., and UK health care systems.CLINICAL IMPLICATIONS OF FFR FROM ECONOMIC PERSPECTIVES.The treatment decision based on FFR has 2 clinical implications.First, FFR provides information to avoid unnecessary PCIs.Second, FFR allows optimal selection of the lesion requiring revascularization.Because the current study exclusively evaluated patients who were treated by PCI, the significantly lower risk of spontaneous MI and allcause death in the FFR-based PCI group was mainly driven by the second clinical role of FFR.In addition, as patients who were deferred PCI based on insignificant FFR was not included in the current study, FFRbased PCIs had higher cost at index admission due to the cost of pressure wire and hyperemic agents.Nevertheless, the additional cost of measuring FFR during the index procedure was offset, and the total cumulative medical costs were lower in the FFR group based on the significantly lower risk of spontaneous MI and all-cause death during 4 years of follow-up.The significantly lower risk of hard clinical outcomes raised QALYs in FFR-based PCIs, consequently, FFR-based PCI was a dominant treatment strategy over angiography-based PCI.
STUDY LIMITATIONS.Some limitations should be acknowledged.First, since the cost-effectiveness analysis was based on a retrospective cohort, susceptibilities related to the study design, such as measured or unmeasured confounding factors, were also inherent.Second, the results of the current study were strongly dependent on key input values such as transition probability, medical costs, and utilities, which were obtained through limited research.However, both NHIS-HIRA data-based and the modelbased analyses showed consistent results, and multiple sensitivity analyses also support the costeffectiveness of FFR-based PCI over angiographybased PCI.Third, the current study assumed common key input values for all individuals, irrespective of their own baseline characteristics, such as age and sex.Fourth, the current results could not be applied to health-care systems other than Korea, U.S., or UK.
Abbott Vascular, Boston Scientific, Philips Volcano, Terumo Corporation, Zoll Medical, and Donga-ST.All other authors have reported that they have no relationships relevant to the contents of this article to disclose.

Table 1
summarizes key input parameters including transition probability, utilities, and medical costs used in the cost-effectiveness analysis.Transition probability for the Korean population was acquired from phase 1. Multivariable Cox proportional hazard regression was used to calculate the adjusted HR and 95% CI to compare the risk of clinical events according to the use of FFR during the index PCI.The transition probability for the U.S. and UK populations was obtained from previous evidence of a nationwide registry (Veterans Affairs Clinical Assessment, Reporting, and Tracking Program) and RCT (FAME [Fractional Flow Reserve versus Angiography for

TABLE 1
Key Inputs in the Model a Adjusted for age, sex, clinical presentation, hypertension, diabetes mellitus, hyperlipidemia, congestive heart failure, previous CVA, atrial fibrillation, peripheral vascular disease, chronic obstructive pulmonary disease, chronic renal failure, type of stent, number of stents, discharge medications, and medical cost during index admission.CVA ¼ cerebrovascular accident; FFR ¼ fractional flow reserve; HIRA ¼ Health Insurance Review and Assessment; MI ¼ myocardial infarction; PCI ¼ percutaneous coronary intervention.

Table 1 ,
To identify the impact of changing the key Supplemental Table2).
Among 134,613 eligible patients who underwent PCI for stable or unstable angina, FFR-based PCI group was used during index PCI in 5,116 patients (3.8%) (Central Illustration upper left panel).The mean age and proportion of female patients were 66.8 AE 10.3 Regarding the effectiveness, there was a gain of 0.039 QALYs in the FFR-based PCI group compared with the angiography-based PCI group during 4 years of follow-up.The total cumulative cost per patient in the FFR-based PCI group was estimated to be $10,503, about $303 less than that of the angiography-based COST-EFFECTIVENESS ANALYSIS USING A DECISION AND MARKOV MODEL.After the end of the 10-year simulation, the cumulative mortality and incident MI rates were 46.4% and 11.7% in the angiographybased PCI group and 38.5% and 9.4% in the FFRbased PCI group, respectively.In base-case analysis, FFR-based PCI in Korea showed better quality of life than angiography-based PCI (4.90 vs 4.60 QALYs) in the model.Simultaneously, FFR-based PCIs resulted in lower medical costs ($21,744 vs $23,951) than angiography-based PCIs (Table5).

TABLE 2
Patient and Procedural Characteristics of Study Population Values are mean AE SD or n (%).ACEI ¼ angiotensin-converting enzyme inhibitor; ARB ¼ angiotensin receptor blocker; CVA ¼ cerebrovascular accident; FFR ¼ fractional flow reserve; NOAC ¼ non-vitamin K antagonist oral anticoagulant; PCI ¼ percutaneous coronary intervention.

TABLE 4
Subgroup Analysis of ICER With FFR-Based PCI Relative to Angiography-Based PCI FFR ¼ fractional flow reserve; ICER ¼ incremental cost-effectiveness ratio; PCI ¼ percutaneous coronary intervention; QALYs ¼ quality-adjusted life years.

TABLE 3
ICER With FFR-Based PCI Relative to Angiography-Based PCI

TABLE 5
10-Cycle Simulation Results of ICER With FFR-Based PCI Relative to Angiography-Based PCI FFR ¼ fractional flow reserve; ICER ¼ incremental cost-effectiveness ratio; PCI ¼ percutaneous coronary intervention; QALYs ¼ quality-adjusted life years.